llama2-ppo: benchmark results
Provider: Meta.
Aggregate Rating 1208 ± 36, rank #9568 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 45.08 | MC2 (%) (0-shot) | 29.6 |
| Open LLM Leaderboard v1 - MMLU | 35.36 | Accuracy (%) (5-shot) | 18.4 |
| Open LLM Leaderboard v1 - WinoGrande | 64.96 | Accuracy (%) (5-shot) | 18 |
| Open LLM Leaderboard v1 - ARC Challenge | 41.64 | Normalized accuracy (%) (25-shot) | 17.8 |
| Open LLM Leaderboard v1 - GSM8K | 0.15 | Accuracy (%) (5-shot) | 10.1 |
| Open LLM Leaderboard v1 - HellaSwag | 49.46 | Normalized accuracy (%) (10-shot) | 10.1 |
Interactive version: theaggregate.ai/model?slug=llama2-ppo · How It Works · Data refreshed daily, snapshot 2026-11-08.